Brazil is launching its first artificial intelligence-focused data center in the Amazon region, powered by 100% renewable energy from hydroelectric, wind, and solar sources. The facility, called BEL1, will open in the second quarter of 2027 with an initial 7.5 megawatts of capacity, scaling to 100 megawatts as demand grows. The project represents a significant bet on whether renewable-powered AI infrastructure can operate responsibly in one of the world's most ecologically critical landscapes.

Elea Data Centers and AXIA Energia are partnering to build BEL1 in Belém, the capital of Pará state in northern Brazil. The location sits on the eastern edge of the Amazon rainforest and watershed, making environmental performance a central concern. According to Elea CEO and founder Alessandro Lombardi, the site was strategically chosen to expand into a historically underserved market and bridge the digital divide in the region.

Why Does Location Matter for AI Data Centers?

The choice of Belém offers practical infrastructure advantages beyond renewable energy. The city provides an alternative route to Fortaleza, Brazil's primary landing point for submarine cables, helping diversify the country's network infrastructure and improve resilience. Belém is also connected to the Northern region's fiber network through the Norte Conectado (Connected North) program and is expected to benefit from the expansion of coastal fiber routes linking South America with Europe.

More importantly, the facility will be built adjacent to AXIA's Miramar high-voltage substation, providing direct access to the transmission network while minimizing transmission and distribution losses. An on-site high-voltage substation and land bank are also planned, allowing the data center to accept high-density demand directly from the grid, maximizing power stability. This infrastructure mirrors the setup used at Elea's São Bernardo do Campo campus.

How Can Data Centers Reduce Their Environmental Footprint?

As AI computing demand accelerates globally, data centers face mounting scrutiny over electricity consumption and water use for cooling. Elea has committed to specific sustainability targets that go beyond simply sourcing renewable power:

• Power Usage Effectiveness (PUE): Elea aims to maintain a PUE lower than 1.5 across all operations. PUE measures how much total facility power is consumed relative to the power used by computing equipment; lower numbers indicate greater efficiency.
• Water Usage Effectiveness (WUE): The company set a 2026 goal of achieving 0.2 liters per kilowatt-hour and plans to reduce it further to 0.18 liters per kilowatt-hour by the end of 2026.
• Certification Standards: BEL1 is designed and constructed to meet sustainability standards, including LEED and ISO certifications, ensuring the facility meets recognized environmental benchmarks.

These metrics matter because data centers consume enormous amounts of water for cooling. By setting aggressive targets, Elea is attempting to demonstrate that large-scale AI infrastructure can operate in water-sensitive regions without depleting local resources.

The Amazon location adds a layer of complexity. While the project is intended to support digital infrastructure and reduce regional inequality, its long-term sustainability will depend not only on renewable energy supply but also on how effectively it manages water use, minimizes impacts on the surrounding ecosystem, and operates alongside one of the world's most ecologically significant landscapes.

What's Driving the Global Push for Renewable-Powered AI Infrastructure?

BEL1 is not an isolated initiative. Across the globe, hyperscalers and data center operators are grappling with the energy demands of artificial intelligence. The U.S. solar industry is expanding rapidly to meet rising electricity demand, with utility-scale solar generation expected to increase 19% this summer compared to last year, driven by a 20% increase in installed solar capacity.

The Energy Information Administration (EIA) forecasts that nearly all incremental electricity demand will be met by renewable energy sources, underscoring solar's expanding role as the backbone of new U.S. power generation. Wind generation is expected to rise 10%, with modest increases in hydroelectric and nuclear output, while coal generation continues to decline.

This shift reflects a broader recognition that renewable energy is becoming one of the lowest-cost and fastest-to-deploy solutions for meeting rising electricity demand driven by economic growth, electrification, and AI-powered data centers. For developers like Elea, sourcing 100% renewable power is increasingly a competitive advantage rather than a constraint.

The investment in BEL1 also signals confidence in Brazil's digital economy. By establishing this facility, Elea aims to significantly strengthen the region's digital infrastructure and reduce digital inequality by improving access to advanced online services outside of Brazil's traditional technology hubs. The expansion is intended to enhance the region's competitiveness within Brazil's growing digital economy.

As AI infrastructure continues to expand globally, projects like BEL1 will serve as a test case for whether large-scale computing facilities can operate sustainably in ecologically sensitive regions. The facility's performance on water management, energy efficiency, and ecosystem impact will likely influence how future AI data centers are sited and operated in the Global South.